I think it must be \theta(x,b) instead of \theta(x,y) at the end of the definition of elimination of imaginaries.